ABC Says Tied (LV) or Bush +1 (RV)

[The Vorlon]

Likely Voters (+/- 56% turnout Model)

Bush 48
Kerry 48
Nader 1

Job Approbal is 52% (Washington Post article says 50%..?)

In the 2 way race, its also tied, 48/48 in LVs, 49/48 Kerry in RVs

Bush leads 53% to 44% in "Battlegrounds" among likely voters

(Sample in Battlegrounds = +/- 320 voters or so +/- 5.6% MOE)

Good news for Bush, but given small sample, don't take it to the bank just yet.

Partisan ID was:

GOP 33%
Dems 32%
Indys 29%
(Refused/other presumably 5%)

If you believe Gallup this is just about perfect, if you believe PEW it is 4% to GOP heavy.
Take your pick

Trend: (August 1st Poll)

Kerry 49
Bush 47
Nader 2

Gain of 2% for Bush, well within MOE

http://www.abcnews.go.com/sections/politics/Vote2004/bush_convention_poll_040829.html

Another poll als showing the "turnout curve" getting very flat:

Registered Voters

Bush 48
Kerry 47
Nader 1%

Trend: (August 1st Poll)

Kerry 50
Bush 44
Nader 2%


By this measure Bush gained 7% which achieves statistical signifigance.

The Skinny...

Horserace numbers little changed...

Internals VERY strong for Bush and very bad for Kerry

In numbers very similar to the Time POll, Kerry's personal favorable/unfavorable took a HUGE hit

Favorable-Unfavorable  (NOW)

George W. Bush 50-40  (Bush +10)
John Kerry 43-40 (Kerry +3)

Favorable-Unfavorable  (August 1)

Bush 47-45 (+2)
Kerry 51-32 (+19)

Kerry went from a 17% advantage on personal favorability to a 7% deficit - a net swing of 24% - This is, from a polling perspective, a fairly staggering number.

[CARLHAYDEN]

As I said before, I think many of the polls produced in July showing Kerry leads were defective.

I suspect that a number of pollsters have, er, 'refined' their methodology to produce more plausible numbers.

I still wonder about Opinion Dynamics move in the opposite direction.

If they are using the same screening method as previously (I believe this is the case), they how is it that there numbers are trending Democrat while all the others are trending Republican?

[Smash255]

This poll has Reps  +1. thats about a +5 too heavy GOP sample (Dems +4 national) if a poll came out +5 the other way +9 on the Dem side, the reps here would be screaming their heads off

[agcatter]

You make a good point.  It is definitely too GOP heavy.  

[ATFFL]

There are those who think the party registration has narrowed to be dead even.  Others think the Dems are up 4.  The truth likely lies inbetween.

My question is are the partisan breakdowns for the RV or LV samples?

[A18]

No one cares about registered voters, anyway. I only take likely voters polls seriously.

[The Vorlon]

This poll has Reps  +1. thats about a +5 too heavy GOP sample (Dems +4 national) if a poll came out +5 the other way +9 on the Dem side, the reps here would be screaming their heads off

As I said..

Partisan ID was:

GOP 33%
Dems 32%
Indys 29%
(Refused/other presumably 5%)

If you believe Gallup this is just about perfect, if you believe PEW it is 4% to GOP heavy.
Take your pick

We are a "full disclosure" poll posting pundit